InfoExchange

You are an intern at a firm and they just say "Buy low, Sell High"! Well, good luck parsing through the computer logs with 2000 transactions/minute...

Comment

InfoExchange

You're an intern on the trading team at your firm. It's your first day and you get assigned to analyze only 4 minutes worth of trading data provided by the stock exchange computers.

So, only 4 minutes? how hard can that be?

Well... To give you an idea, there are over 2000 transaction logs done by a typical firm (e.g. National Bank) per minute, and that is only a fraction of everything that might happen on a typical day. So, without the analytics covered by the InfoExchange dashboard, this internship might be harder than you anticipated :,)

Inspired by National Bank Challenge At ConuHacksVIII.

What it does

  • Identify trends, anomalies and patterns across all trades done in Exchanges
  • Ability to expand, zoom in/out, drag, and dynamically interact with graphs
  • Graphs about stock trading prices, offer prices and cancelation rate
  • Functionality to save graphs as you interact with the trends and patterns
  • Coded dynamically, any other data can be uploaded and displayed in the data dashboard

How we built it

Credit: Logo created by Vista

Python data libraries: python for json parsing, streamlit, and graphing libraries (plotly, seahorn, bokeh)

Data provided by National Bank of Canada

Challenges we ran into

A lot of transactions and logs! Was tough to narrow it down at first and thinking of ways to parse the data effectively was a challenge.

What we learned

Proud of learning python plotting tools and had good practice parsing through all the json data. Brushed up on my basic python skills.

What's next for InfoExchange

  • AI-integeration to analyze the graphs and identify anomalies
  • Add more variety of graphs that might be useful

Built With

Share this project:

Updates